Codebook: Essex Dual Process (With Omnibus 2019)

codebook(sx_dp, survey_repetition = "single", metadata_table = TRUE)

knitr::asis_output(data_info)

Metadata

Description

if (exists("name", meta)) {
  glue::glue(
    "__Dataset name__: {name}",
    .envir = meta)
}

Dataset name: Dual process, effectiveness and empathy in charitable giving decisions

cat(description)

Essex 2019 - Omnibus survey, Charitable asks for blindness charities with effectiveness and empathy treatments, ESSExLab pool: Non-British and late responders

• Temporal Coverage: 2019
• Spatial Coverage: Essex, UK
• Citation: Bergh, Reinstein (2019)
• URL: TBA
• Identifier: doi:TBA

• Date published: TBA

• Creator:

  • • @type: Person
    • givenName: David
    • familyName: Reinstein
    • email: daaronr@gmail.com

    • affiliation:

        * **@type**: Organization
        * **name**: University of Exeter

  • • @type: Person
    • givenName: Robin
    • familyName: Bergh
    • email:

    • affiliation:

        * **@type**: Organization
        * **name**: Harvard University

meta <- meta[setdiff(names(meta),
                     c("creator", "datePublished", "identifier",
                       "url", "citation", "spatialCoverage", 
                       "temporalCoverage", "description", "name"))]
pander::pander(meta)

• keywords: Experiment, Charitable giving, Survey, Empathy, Happiness, Dual process, Effectiveness, Empathy and Identifiable victim

knitr::asis_output(survey_overview)

Variables

knitr::asis_output(paste0(scales_items, sep = "\n\n\n", collapse = "\n\n\n"))

dad_job

What is your father’s occupation? If he is retired or deceased, what was his occupation before retirement?

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

2 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	label	data_type	missing	complete	n	empty	n_unique	min	max
dad_job	What is your father’s occupation? If he is retired or deceased, what was his occupation before retirement?	character	2	376	378	0	10	5	47

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

decision_mode

Think of when you make a decision. Generally speaking, do you tend to decide rather quickly relying mostly on your intuition or rather do you tend to think accurately about all possible alternatives and consequences of your choice, taking as much time as needed before reaching a final decision?

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	label	data_type	missing	complete	n	empty	n_unique	min	max
decision_mode	Think of when you make a decision. Generally speaking, do you tend to decide rather quickly relying mostly on your intuition or rather do you tend to think accurately about all possible alternatives and consequences of your choice, taking as much time as needed before reaching a final decision?	character	0	378	378	0	4	32	56

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

impairment

Please describe your impairment.

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 36  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

340 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	label	data_type	missing	complete	n	empty	n_unique	min	max
impairment	Please describe your impairment.	character	340	38	378	0	36	5	110

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

mom_job

What is your mother’s occupation? If she is retired or deceased, what was her occupation before retirement?

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

2 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	label	data_type	missing	complete	n	empty	n_unique	min	max
mom_job	What is your mother’s occupation? If she is retired or deceased, what was her occupation before retirement?	character	2	376	378	0	10	5	47

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

years_in_uk

How many years have you lived in the United Kingdom?

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

116 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	label	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
years_in_uk	How many years have you lived in the United Kingdom?	numeric	116	262	378	5.17	8.1	0	1	3	5	61	▇▁▁▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

start_date

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 378  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	n_unique	median	min	max
start_date	POSIXct	0	378	378	378	2019-06-03	2019-05-31	2019-06-10

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

end_date

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 378  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	n_unique	median	min	max
end_date	POSIXct	0	378	378	378	2019-06-03	2019-05-31	2019-06-10

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

progress

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
progress	numeric	0	378	378	99.95	0.5	91	100	100	100	100	▁▁▁▁▁▁▁▇

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

duration_in_seconds

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
duration_in_seconds	numeric	0	378	378	24351.7	83162.28	304	942.5	1242	2175.75	605522	▇▁▁▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

finished

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	count	mean
finished	logical	0	378	378	TRU: 370, FAL: 8, NA: 0	0.98

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

recorded_date

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 377  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	n_unique	median	min	max
recorded_date	POSIXct	0	378	378	377	2019-06-03	2019-05-31	2019-06-11

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

response_id

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 378  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
response_id	character	0	378	378	0	378	17	17

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_1	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_2	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_3	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_4

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_4	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_5

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_5	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_6

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_6	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_7

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_7	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_8

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_8	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_9

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_9	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attributes_10

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attributes_10	character	0	378	378	0	7	14	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_1_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_1_1	character	0	378	378	0	6	7	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_1_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_1_2	character	0	378	378	0	6	7	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_1_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_1_3	character	0	378	378	0	5	7	17

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_1_4

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_1_4	character	0	378	378	0	6	7	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_1_5

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_1_5	character	0	378	378	0	6	7	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_2_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_2_1	character	0	378	378	0	6	7	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_2_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_2_2	character	0	378	378	0	6	7	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_2_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_2_3	character	0	378	378	0	6	7	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_2_4

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_2_4	character	0	378	378	0	6	7	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

schwartz_s_2_5

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
schwartz_s_2_5	character	0	378	378	0	6	7	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

auth_s_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
auth_s_1	character	0	378	378	0	2	12	24

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

auth_s_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
auth_s_2	character	0	378	378	0	2	9	12

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

auth_s_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
auth_s_3	character	0	378	378	0	2	9	13

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

auth_s_4

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
auth_s_4	character	0	378	378	0	2	17	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_1_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_1_1	character	0	378	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_1_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_1_2	character	0	378	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_1_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_1_3	character	0	378	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_1_4

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_1_4	character	0	378	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_1_5

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_1_5	character	0	378	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_1_6

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_1_6	character	1	377	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_2_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_2_1	character	0	378	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_2_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_2_2	character	0	378	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_2_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_2_3	character	0	378	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_2_4

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_2_4	character	1	377	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_2_5

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_2_5	character	0	378	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

topics_s_2_6

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
topics_s_2_6	character	1	377	378	0	5	5	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

worth

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
worth	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	0	378	378	4	Agr: 188, Str: 132, Dis: 49, Str: 9

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

good_qualties

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
good_qualties	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	0	378	378	4	Agr: 223, Str: 132, Dis: 20, Str: 3

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

failure

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
failure	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	0	378	378	4	Dis: 168, Str: 134, Agr: 61, Str: 15

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

comparison

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
comparison	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	0	378	378	4	Agr: 232, Str: 102, Dis: 36, Str: 8

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

proud

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
proud	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	0	378	378	4	Dis: 171, Str: 103, Agr: 85, Str: 19

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

positive_attitude

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
positive_attitude	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	0	378	378	4	Agr: 197, Str: 87, Dis: 76, Str: 18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

satisfaction

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
satisfaction	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	0	378	378	4	Agr: 216, Dis: 82, Str: 70, Str: 10

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

respect

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
respect	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	0	378	378	4	Agr: 130, Dis: 125, Str: 71, Str: 52

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

useless

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
useless	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	0	378	378	4	Agr: 165, Dis: 108, Str: 59, Str: 46

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

no_good

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
no_good	character	0	378	378	0	4	5	17

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

risk

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
risk	factor	TRUE	1. Completely unwilling to take risks, 2. 1, 3. 2, 4. 3, 5. 4, 6. Risk Neutral, 7. 6, 8. 7, 9. 8, 10. 9, 11. Very willing to take risks	0	378	378	11	7: 87, 6: 70, 8: 60, Ris: 53

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

trust

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
trust	character	0	378	378	0	3	7	26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

decision_mode_follow

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

192 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
decision_mode_follow	character	192	186	378	0	3	32	35

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

likert_time_recip_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
likert_time_recip_1	character	0	378	378	0	11	1	29

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

likert_time_recip_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
likert_time_recip_2	character	0	378	378	0	11	1	29

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

likert_time_recip_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
likert_time_recip_3	character	0	378	378	0	11	1	29

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

pol_interest

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
pol_interest	factor	TRUE	1. Very interested, 2. Somewhat interested, 3. Not very interested, 4. Not at all interested	0	378	378	4	Som: 138, Not: 102, Ver: 90, Not: 48

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

political

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
political	factor	TRUE	1. Far Left, 2. 2, 3. 3, 4. 4, 5. 5, 6. 6, 7. 7, 8. 8, 9. 9, 10. Far Right	1	377	378	10	5: 112, 3: 64, 4: 56, 6: 48

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

social_ideo_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
social_ideo_1	factor	TRUE	1. Never justifiable, 2. 2, 3. 3, 4. 4, 5. 5, 6. 6, 7. 7, 8. 8, 9. 9, 10. Always Justifiable	1	377	378	10	Alw: 231, 5: 31, Nev: 30, 9: 22

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

social_ideo_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
social_ideo_2	factor	TRUE	1. Never justifiable, 2. 2, 3. 3, 4. 4, 5. 5, 6. 6, 7. 7, 8. 8, 9. 9, 10. Always Justifiable	1	377	378	10	Alw: 69, 5: 62, 7: 51, Nev: 45

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

social_ideo_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
social_ideo_3	factor	TRUE	1. Never justifiable, 2. 2, 3. 3, 4. 4, 5. 5, 6. 6, 7. 7, 8. 8, 9. 9, 10. Always Justifiable	1	377	378	10	Alw: 162, 5: 40, 7: 38, 8: 32

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

social_ideo_4

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
social_ideo_4	factor	TRUE	1. Never justifiable, 2. 2, 3. 3, 4. 4, 5. 5, 6. 6, 7. 7, 8. 8, 9. 9, 10. Always Justifiable	1	377	378	10	Alw: 184, 8: 37, 5: 36, 9: 36

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

social_ideo_5

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
social_ideo_5	factor	TRUE	1. Never justifiable, 2. 2, 3. 3, 4. 4, 5. 5, 6. 6, 7. 7, 8. 8, 9. 9, 10. Always Justifiable	1	377	378	10	Alw: 222, 9: 31, Nev: 22, 5: 22

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

ineq

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## No non-missing values to show.

knitr::opts_chunk$set(fig.height = old_height)

378 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
ineq	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	378	0	378	0	NA: 378, Str: 0, Dis: 0, Agr: 0

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

redist

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## No non-missing values to show.

knitr::opts_chunk$set(fig.height = old_height)

378 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
redist	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	378	0	378	0	NA: 378, Str: 0, Dis: 0, Agr: 0

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

standard_live

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## No non-missing values to show.

knitr::opts_chunk$set(fig.height = old_height)

378 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
standard_live	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	378	0	378	0	NA: 378, Str: 0, Dis: 0, Agr: 0

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

private_redist

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## No non-missing values to show.

knitr::opts_chunk$set(fig.height = old_height)

378 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
private_redist	factor	TRUE	1. Strongly Disagree, 2. Disagree, 3. Agree, 4. Strongly Agree	378	0	378	0	NA: 378, Str: 0, Dis: 0, Agr: 0

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

tax_share

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

13 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
tax_share	factor	TRUE	1. Much smaller share, 2. Smaller, 3. The same share, 4. Larger, 5. Much larger share	13	365	378	5	Lar: 205, The: 93, Muc: 58, NA: 13

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

uk_aid_spend

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
uk_aid_spend	factor	TRUE	1. Spend less, 2. Spend the same amount, 3. Spend more	0	378	378	3	Spe: 187, Spe: 134, Spe: 57, NA: 0

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

uk_immig_lvl

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
uk_immig_lvl	character	0	378	378	0	3	29	35

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

uk_poverty_aid

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

187 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
uk_poverty_aid	character	187	191	378	0	7	15	21

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

uk_problems

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

187 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
uk_problems	character	187	191	378	0	5	5	17

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

uk_similar

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

187 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
uk_similar	character	187	191	378	0	6	12	21

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

188 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_1	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	188	190	378	5	NA: 188, Ver: 60, Mod: 53, Ext: 50

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

188 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_2	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	188	190	378	5	NA: 188, Ver: 67, Mod: 55, Ext: 44

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

188 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_3	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	188	190	378	5	NA: 188, Ext: 91, Ver: 64, Mod: 28

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_4

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

188 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_4	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	188	190	378	5	NA: 188, Ext: 103, Ver: 47, Mod: 25

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_5

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

188 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_5	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	188	190	378	5	NA: 188, Ext: 88, Ver: 73, Mod: 26

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_6

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

189 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_6	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	189	189	378	5	NA: 189, Ext: 78, Ver: 55, Mod: 38

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_7

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

188 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_7	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	188	190	378	5	NA: 188, Ver: 61, Mod: 59, Ext: 46

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_8

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

188 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_8	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	188	190	378	5	NA: 188, Mod: 63, Ver: 49, Sli: 35

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_9

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

188 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_9	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	188	190	378	5	NA: 188, Not: 55, Mod: 50, Sli: 32

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_uk_10

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

188 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	ordered	value_labels	missing	complete	n	n_unique	top_counts
get_ahead_uk_10	factor	TRUE	1. Not at all important, 2. Slightly important, 3. Moderately important, 4. Very important, 5. Extremely important	188	190	378	5	NA: 188, Not: 52, Mod: 47, Ver: 39

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

sex

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
sex	character	0	378	378	0	3	4	6

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

birth

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 356  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

4 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
birth	character	4	374	378	0	356	10	10

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

race

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
race	character	0	378	378	0	8	4	30

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

religion

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
religion	character	0	378	378	0	7	5	9

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

single

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
single	character	1	377	378	0	3	6	18

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

uni_student

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
uni_student	character	1	377	378	0	2	2	3

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

studies

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 175  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

119 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
studies	character	119	259	378	0	175	2	64

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

prev_part

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
prev_part	character	0	378	378	0	2	2	3

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

exp_part

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
exp_part	character	0	378	378	0	3	11	46

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

impaired

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
impaired	character	0	378	378	0	2	2	3

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

mom_education

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

3 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
mom_education	character	3	375	378	0	6	15	82

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

dad_education

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

2 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
dad_education	character	2	376	378	0	6	15	82

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

siblings

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

2 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
siblings	numeric	2	376	378	1.69	1.41	0	1	1	2	8	▇▃▂▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

siblings_order

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

64 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
siblings_order	numeric	64	314	378	0.85	1.03	0	0	1	1	6	▇▆▂▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

children

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

2 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
children	character	2	376	378	0	2	2	3

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

number_children

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

310 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
number_children	numeric	310	68	378	1.91	0.88	1	1	2	2	5	▆▇▁▃▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

birth_nation

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

2 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
birth_nation	character	2	376	378	0	2	2	3

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

abroad

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

2 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
abroad	character	2	376	378	0	2	2	3

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

nation

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 82  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

116 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
nation	character	116	262	378	0	82	2	24

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

first_language

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

2 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
first_language	character	2	376	378	0	2	2	3

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

home_language

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 74  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

169 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
home_language	character	169	209	378	0	74	1	25

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

languages

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 183  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

40 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
languages	character	40	338	378	0	183	1	83

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

other_country_s_1_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 97  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

138 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
other_country_s_1_1	character	138	240	378	0	97	1	27

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

other_country_s_1_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 67  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

218 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
other_country_s_1_2	character	218	160	378	0	67	1	24

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

other_country_s_1_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 40  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

299 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
other_country_s_1_3	character	299	79	378	0	40	1	24

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

other_country_s_2_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 83  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

143 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
other_country_s_2_1	character	143	235	378	0	83	1	21

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

other_country_s_2_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 59  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

220 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
other_country_s_2_2	character	220	158	378	0	59	1	25

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

other_country_s_2_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## 32  unique, categorical values, so not shown.

knitr::opts_chunk$set(fig.height = old_height)

301 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
other_country_s_2_3	character	301	77	378	0	32	1	13

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

dev_poverty_aid

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
dev_poverty_aid	character	191	187	378	0	7	15	21

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

dev_problems

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
dev_problems	character	191	187	378	0	5	5	17

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

dev_similar

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
dev_similar	character	191	187	378	0	6	12	21

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_1

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_1	character	191	187	378	0	5	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_2

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_2	character	191	187	378	0	5	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_3

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_3	character	191	187	378	0	4	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_4

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_4	character	191	187	378	0	5	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_5

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_5	character	191	187	378	0	5	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_6

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_6	character	191	187	378	0	5	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_7

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

192 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_7	character	192	186	378	0	5	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_8

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_8	character	191	187	378	0	5	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_9

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

192 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_9	character	192	186	378	0	5	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

get_ahead_dev_10

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

191 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
get_ahead_dev_10	character	191	187	378	0	5	14	20

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

charchc_noeff_6

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

356 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
charchc_noeff_6	character	356	22	378	0	1	101	101

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

charchc_noeff_7

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

353 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
charchc_noeff_7	character	353	25	378	0	1	96	96

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

charch_noef_rev_7

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

339 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
charch_noef_rev_7	character	339	39	378	0	1	97	97

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

charch_noef_rev_12

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

360 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
charch_noef_rev_12	character	360	18	378	0	1	98	98

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

char_choice_efftreat_6

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

357 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
char_choice_efftreat_6	character	357	21	378	0	1	98	98

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

char_choice_efftreat_7

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

343 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
char_choice_efftreat_7	character	343	35	378	0	1	96	96

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

choice_eff_reverse_6

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}

## Error in as.environment(where): using 'as.environment(NULL)' is defunct

attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

348 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
choice_eff_reverse_6	character	348	30	378	0	1	96	96

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

choice_eff_reverse_12

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

359 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
choice_eff_reverse_12	character	359	19	378	0	1	98	98

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

donation

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
donation	numeric	0	378	378	9.33	13.44	0	0	2	15	50	▇▂▁▂▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

attn_geog_qn

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

34 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
attn_geog_qn	character	34	344	378	0	6	6	24

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

reaction_qn

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

26 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
reaction_qn	character	26	352	378	0	2	30	39

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

happiness

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

9 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
happiness	character	9	369	378	0	7	14	25

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

aid_treat

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
aid_treat	character	0	378	378	0	2	2	3

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

treat_image

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
treat_image	character	0	378	378	0	2	5	8

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

treat_eff_info

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
treat_eff_info	character	0	378	378	0	2	4	7

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

d_river_1_st

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	count	mean
d_river_1_st	logical	0	378	378	FAL: 191, TRU: 187, NA: 0	0.49

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

time_desc_chars

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

189 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
time_desc_chars	numeric	189	189	378	25.03	25.61	4.21	11.53	17.98	28.54	225.23	▇▂▁▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

time_charch

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

1 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
time_charch	numeric	1	377	378	22.52	20.47	3.12	11.11	16.61	28.09	225.23	▇▂▁▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

time_intro_prize

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
time_intro_prize	numeric	0	378	378	24.03	20.58	1.11	11.29	19.51	30.23	174.98	▇▅▁▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

treatment

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
treatment	character	0	378	378	0	4	7	10

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

char_choice

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
char_choice	character	0	378	378	0	3	10	15

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

d_donation

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
d_donation	character	0	378	378	0	2	1	2

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

d_donated_all

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
d_donated_all	numeric	0	378	378	0.05	0.22	0	0	0	0	1	▇▁▁▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

d_donated_mid

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
d_donated_mid	numeric	0	378	378	0.46	0.5	0	0	0	1	1	▇▁▁▁▁▁▁▇

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

endow

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
endow	numeric	0	378	378	50	0	50	50	50	50	50	▁▁▁▇▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

av_don

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
av_don	numeric	0	378	378	0.19	0.27	0	0	0.04	0.3	1	▇▂▁▂▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

don_guide_dogs

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
don_guide_dogs	numeric	0	378	378	3.13	8.7	0	0	0	0	50	▇▁▁▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

d_don_guide_dogs

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
d_don_guide_dogs	numeric	0	378	378	0.19	0.4	0	0	0	0	1	▇▁▁▁▁▁▁▂

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

don_river

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
don_river	numeric	0	378	378	6.19	12	0	0	0	7	50	▇▁▁▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

d_don_river

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
d_don_river	numeric	0	378	378	0.33	0.47	0	0	0	1	1	▇▁▁▁▁▁▁▃

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

date_sent

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	n_unique	median	min	max
date_sent	POSIXct	0	378	378	1	2019-05-31	2019-05-31	2019-05-31

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

dur_survey

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
dur_survey	numeric	0	378	378	405.87	1386.04	5.07	15.71	20.71	36.27	10092.03	▇▁▁▁▁▁▁▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

delay_response

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
delay_response	numeric	0	378	378	2.76	2.83	0.0019	0.18	2.59	3.9	10.1	▇▁▅▁▂▁▁▂

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

eyes_correct_pct

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

2 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
eyes_correct_pct	numeric	2	376	378	0.69	0.14	0.17	0.61	0.72	0.78	1	▁▁▁▃▇▇▅▁

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

d_decis_intu

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
d_decis_intu	numeric	0	378	378	0.34	0.47	0	0	0	1	1	▇▁▁▁▁▁▁▅

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

d_decis_intu_m

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	mean	sd	p0	p25	p50	p75	p100	hist
d_decis_intu_m	numeric	0	378	378	0.34	0.47	0	0	0	1	1	▇▁▁▁▁▁▁▅

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

dow

Distribution

show_missing_values <- FALSE
if (has_labels(item)) {
  missing_values <- item[is.na(haven::zap_missing(item))]
  attributes(missing_values) <- attributes(item)
  if (!is.null(attributes(item)$labels)) {
    attributes(missing_values)$labels <- attributes(missing_values)$labels[is.na(attributes(missing_values)$labels)]
    attributes(item)$labels <- attributes(item)$labels[!is.na(attributes(item)$labels)]
  }
  if (is.double(item)) {
    show_missing_values <- length(unique(haven::na_tag(missing_values))) > 1
    item <- haven::zap_missing(item)
  }
  if (length(item_attributes$labels) == 0 && is.numeric(item)) {
    item <- haven::zap_labels(item)
  }
}
item_nomiss <- item[!is.na(item)]

# unnest mc_multiple and so on
if (
  is.character(item_nomiss) &&
  any(stringr::str_detect(item_nomiss, stringr::fixed(", "))) &&
  (exists("type", item_info) && 
    any(stringr::str_detect(item_info$type, 
                            pattern = stringr::fixed("multiple"))))
  ) {
  item_nomiss <- unlist(stringr::str_split(item_nomiss, pattern = stringr::fixed(", ")))
}
attributes(item_nomiss) <- attributes(item)

old_height <- knitr::opts_chunk$get("fig.height")
non_missing_choices <- item_attributes[["labels"]]
many_labels <- length(non_missing_choices) > 7
go_vertical <- !is.numeric(item_nomiss) || many_labels
if ( go_vertical ) {
  # numeric items are plotted horizontally (because that's what usually expected)
  # categorical items are plotted vertically because we can use the screen real estate better this way

    if (is.null(choices) || 
        dplyr::n_distinct(item_nomiss) > length(non_missing_choices)) {
        non_missing_choices <- unique(item_nomiss)
        names(non_missing_choices) <- non_missing_choices
    }
  choice_multiplier <- old_height/6.5
    new_height <- 2 + choice_multiplier * length(non_missing_choices)
    new_height <- ifelse(new_height > 20, 20, new_height)
    new_height <- ifelse(new_height < 1, 1, new_height)
    knitr::opts_chunk$set(fig.height = new_height)
}

wrap_at <- knitr::opts_chunk$get("fig.width") * 10

# todo: if there are free-text choices mingled in with the pre-defined ones, don't show
# todo: show rare items if they are pre-defined
# todo: bin rare responses into "other category"
if (!length(item_nomiss)) {
  cat("No non-missing values to show.")
} else if (is.numeric(item_nomiss) || dplyr::n_distinct(item_nomiss) < 20) {
  plot_labelled(item_nomiss, item_name, wrap_at, go_vertical)
} else {
    cat(dplyr::n_distinct(item_nomiss), " unique, categorical values, so not shown.")
}

knitr::opts_chunk$set(fig.height = old_height)

0 missing values.

Summary statistics

attributes(item) <- item_attributes
df = data.frame(item, stringsAsFactors = FALSE)
names(df) = html_item_name
escaped_table(codebook_table(df))

name	data_type	missing	complete	n	empty	n_unique	min	max
dow	character	0	378	378	0	7	1	1

if (show_missing_values) {
  plot_labelled(missing_values, item_name, wrap_at)
}

if (!is.null(item_info)) {
  # don't show choices again, if they're basically same thing as value labels
  if (!is.null(choices) && !is.null(item_info$choices) && 
    all(names(na.omit(choices)) == item_info$choices) &&
    all(na.omit(choices) == names(item_info$choices))) {
    item_info$choices <- NULL
  }
  item_info$label_parsed <- 
    item_info$choice_list <- item_info$study_id <- item_info$id <- NULL
  pander::pander(item_info)
}

if (!is.null(choices) && length(choices) && length(choices) < 30) {
    pander::pander(as.list(choices))
}

missingness_report

Missingness report

Among those who finished the survey. Only variables that have missing values are shown.

if (!exists("ended", results) ||
  !exists("expired", results)) {
  warning("Could not figure out who finished the surveys, because the ",
          "variables expired and ended were missing.")
}
if (length(md_pattern)) {
  if (knitr::is_html_output()) {
    rmarkdown::paged_table(md_pattern, options = list(rows.print = 10))
  } else {
    knitr::kable(md_pattern)
  }
}

items

Codebook table

export_table(metadata_table)

jsonld